Why semaglutide alone beat the CRV431 combo

A new mouse study asked whether pairing semaglutide with an experimental liver drug called CRV431 would slow liver cancer better than semaglutide on its own. It did not. Mice that got semaglutide by itself developed far fewer liver tumors than mice given the combination, and none of the treatments reduced liver scarring in this advanced-disease model^[1]. The result comes from mice, not people. CRV431 has no approved use anywhere, and this study does not establish that either compound treats liver disease in humans.

What the researchers tested

Researchers at Scripps Research in La Jolla, California, published the work in PLOS One in September 2026^[1]. They used male mice on a Western-style diet with sugar water, plus repeated low doses of a liver toxin (carbon tetrachloride), for 10 weeks. That combination reliably produces the scarring and early tumor nodules seen in metabolic dysfunction-associated steatotic liver disease, or MASLD, the newer clinical name for what used to be called fatty liver disease. By week 10 the mice already had established fibrosis and precancerous changes before any drug was given.

From that starting point, the mice split into groups for eight more weeks of treatment: an untreated group, semaglutide alone (an injected dose raised gradually to a steady maintenance level), CRV431 alone (given by mouth daily), and the two together. Semaglutide, sold as Wegovy for weight management^[4], is a GLP-1 receptor agonist already approved and prescribed in people. CRV431 is not. It is an investigational cyclophilin inhibitor, a drug that blocks a family of proteins some viruses and cancer cells depend on, and it has reached only small early human safety testing so far^[3].

What semaglutide alone did

On tumor burden, semaglutide alone worked best. Researchers scored liver tumor nodules on a 0-to-7 scale. Mice on semaglutide alone averaged 0.44, against 2.89 in the untreated group^[1]. That gap was large enough to count as statistically significant. Semaglutide-treated mice also lost noticeably more body weight over the eight weeks than any other group, a steady decline instead of a brief early dip. CRV431 by itself did something too, just less: its average tumor score was 1.89. Head to head, semaglutide alone outperformed CRV431 alone on tumor burden.

Why the combination did not add up

Here is the surprising part. Mice given semaglutide and CRV431 together averaged a tumor score of 2.33, close to the untreated group and far worse than semaglutide produced on its own^[1]. Adding the second drug did not stack an extra benefit on top of semaglutide's effect. It mostly erased it. That runs against what an earlier study of CRV431 alone, in a different chronic liver-disease model, had suggested: dosed at an advanced disease stage, CRV431 alone cut tumor number and size by roughly half^[2]. Two compounds that each showed promise separately produced a worse tumor result together than semaglutide managed on its own.

The fibrosis side of the story stayed flat across every group. Liver scarring, measured by collagen staining and a direct chemical marker called hydroxyproline, did not differ in any statistically meaningful way between the untreated, semaglutide, CRV431, or combination groups^[1]. Blood markers of liver-cell injury also failed to separate the groups. Once fibrosis and precancerous change are already established, this particular pairing, at this particular dose and duration, was not enough to reverse the scarring, whatever it did to tumor counts.

What this means for people

None of this changes what is approved for people today. Semaglutide is authorized in the EU as Wegovy for weight management in adults, and in some markets for type-2 diabetes. It is not authorized for liver fibrosis or liver cancer^[4]. CRV431, also known by its clinical name rencofilstat, has not been approved anywhere. Its furthest step into human testing so far is a small Phase 1 study checking how the body absorbs a single dose in 44 healthy volunteers, not a trial in people who have liver disease^[3]. Reading this mouse result as evidence that either compound treats human liver cancer would be getting well ahead of the data.

The literature reports that liver-disease findings in mice translate to people unevenly at best; mouse and human livers differ in drug metabolism, immune response, and how fibrosis actually progresses. Nothing here is a reason to seek out CRV431 outside a formal clinical trial, and nothing here changes how semaglutide is prescribed. If you are considering either compound for any reason, that conversation belongs with a clinician who knows your history.

What we don't yet know

This was one model, one sex of mouse, and one starting point: an advanced stage of disease with fibrosis and tumor nodules already present before treatment began. Whether starting the combination earlier, adjusting either dose, or treating for longer would change the picture is genuinely unclear, and the paper does not test it. Nor does it explain the biological reason CRV431 blunted semaglutide's effect on tumors instead of adding to it. Answering that would take further animal work before anyone could responsibly design a human trial pairing the two. For now, the honest summary is interesting, unresolved, and years from any clinical use.